OBJECTIVES: To compare antidiabetic treatment compliance with regards to the fixed or free treatments combination. METHODS: A retrospective study based on the IMS LifeLinkTM Diabetes Cohort linked to IMS Disease Analyzer, a longitudinal patient database containing the electronic health records of patients followed-up by a representative panel of French general practitioners. Two cohorts of patients treated both with metformin and DPPIV were constituted, the first one treated with fixed associations (“fixed cohort”), the second one treated with free associations (“free cohort”). Study investigated patients’ demographic and clinical characteristics, conditions of use, persistence rates, medication possession ratio (MPR) and proportion of days covered (PDC) of the two cohorts. RESULTS: Respectively 2 234 and 8220 patients have been treated with free and fixed antidiabetic associations between 2007 and 2013. Patients are significantly older (66 versus 64 years, p<0.01) in the free cohort where the proportion of men is significantly lower (56% versus 63%, p<0.0001). Nevertheless diabetes oldness is the same (about 10 years) in both cohorts.  Daily number of prescribed tablets whatever the therapeutic classes is significantly higher in the free cohort (4.4 per day versus 3.1, p<0.01). At the end of the first treatment year, significantly more patients of the fixed cohort are still treated by the same association regardless daily dosages or potential add-on treatments (85% versus 72%, p<0.0001); moreover MPR is the same in both free and fixed cohort (64% and 66% without statistical difference) while PDC is significantly higher in the fixed cohort (60% versus 54%, P<0.0001). CONCLUSIONS: Patients treated by fixed antidiabetic treatments associations seem to be more persistent than these treated by free ones at the end of the first year of treatment. These results will be consolidated soon by comparing two similar sub-cohorts with the same medical profile.